A practical hack to improve AI agent reliability is to avoid built-in tool-calling functions. LLMs have more training data on writing code than on specific tool-use APIs. Prompting the agent to write and execute the code that calls a tool leverages its core strength and produces better outcomes.

The significant leap in LLMs isn't just better text generation, but their ability to autonomously execute complex, sequential tasks. This 'agentic behavior' allows them to handle multi-step processes like scientific validation workflows, a capability earlier models lacked, moving them beyond single-command execution.

Instead of interacting with a single LLM, users will increasingly call an API that represents a "system as a model." Behind the scenes, this triggers a complex orchestration of multiple specialized models, sub-agents, and tools to complete a task, while maintaining a simple user experience.

An LLM shouldn't do math internally any more than a human would. The most intelligent AI systems will be those that know when to call specialized, reliable tools—like a Python interpreter or a search API—instead of attempting to internalize every capability from first principles.

The LLM itself only creates the opportunity for agentic behavior. The actual business value is unlocked when an agent is given runtime access to high-value data and tools, allowing it to perform actions and complete tasks. Without this runtime context, agents are merely sophisticated Q&A bots querying old data.

Conceptualize Large Language Models as capable interns. They excel at tasks that can be explained in 10-20 seconds but lack the context and planning ability for complex projects. The key constraint is whether you can clearly articulate the request to yourself and then to the machine.

An AI coding agent's performance is driven more by its "harness"—the system for prompting, tool access, and context management—than the underlying foundation model. This orchestration layer is where products create their unique value and where the most critical engineering work lies.

The LAM is not a model in the traditional sense, but an agent system. It uses the best available LLMs for language understanding and connects them to Rabbit's proprietary tech for controlling actions, allowing for modular upgrades of the underlying AI.

As AI models execute tasks via function calling, their internal state is insufficient for reliable, repeatable business outcomes. They must integrate with external systems (like BPMS) to become predictable "runtimes," ensuring consistent results despite prompt failures or hallucinations.